EPD Comparability Rules: Ensure Apples Match Apples
Two Environmental Product Declarations may look alike at first glance, yet comparing them can be as futile as judging a grapefruit against a granny-smith. Misreading the fine print risks bad purchasing calls, lost tenders, and greenwashing claims that boomerang in courtrooms. This quick-fire guide shows when EPDs *are* truly comparable, and the red flags that say "don’t even try."
Why comparability isn’t baked in
Type III EPDs follow ISO 14025, but the standard itself admits comparisons are valid only under very specific conditions (EPD International, 2024). Two sheets pulled from different program operators, or even two versions of the same Product Category Rule (PCR), can tell different stories.
Same PCR, same game
Start by checking that both declarations cite the identical PCR ID and version. A carpet EPD based on PCR 2019:14 cannot be lined up with one stamped PCR 2024:02 because allocation rules, cut-offs, and scenario defaults shift between versions. If the PCRs mismatch, stop comparing and grab popcorn instead; the numbers are entertainment, not evidence.
Functional unit: the measuring cup
EN 15804 demands a declared or functional unit that ties results to one clear performance description—think "1 m² installed panel for 60 years.” If one EPD uses a declared unit (mass) and the other a functional unit (service life), normalisation is possible but messy and error-prone. When the units diverge, extra math invites accidental rounding errors that hide tonnes of CO₂.
System boundaries and LCA modules
Full cradle-to-grave coverage (A1–C4 plus D) versus cradle-to-gate totals are apples and oranges. ISO 14025 says the boundary must be identical or demonstrably equivalent for comparisons. For construction products, EN 15804 section 5.3 reinforces the rule: partial EPDs may inform building LCAs but cannot act as head-to-head scorecards (ISO 14025, 2024).
Impact indicators and versions
A1-era EPDs use classic CML or TRACI metrics. A2-era EPDs split climate impacts into fossil, biogenic, and land-use buckets. Converting one to the other is like translating slang; nuance gets lost. The Dutch National Environmental Database now rejects any A1 datasets submitted after 1 July 2025, forcing manufacturers onto A2 to keep data road-worthy (NMD, 2025).
Data quality and age
Most program rules cap underlying life-cycle inventory data at ten years. An EPD issued in 2018 may still be technically valid, yet its background electricity mix can differ wildly from a 2025 dataset. Older numbers inflate or deflate scores by double-digit percentages in energy-intensive sectors like glass and steel. If publication dates vary by more than five years, tread carefully.
Verification level and program operator handshake
Third-party verification is mandatory, but schemes vary. Mutual Recognition Agreements (MRAs) between operators like EPD International and PEP Ecopassport broaden the comparison playing field by aligning PCR libraries across borders (EPD International, 2025). Always check whether the two EPDs sit under operators with an active MRA; if not, deeper due-diligence is required.
Quick checklist before you hit “compare”
- PCR ID and version identical?
- Same functional or declared unit?
- Matching system boundaries and LCA modules?
- Impact categories calculated with the same standard version (A2 vs A1)?
- Publication dates within five years and datasets of similar vintage?
- Verified under operators that share an MRA or equivalency deal?
If any answer is "no," the EPDs does'nt pass the sniff test for a fair comparison.
Final thought
EPD comparability is less about the glossy front page and more about the footnotes: PCR, functional unit, boundaries, data age, and verification pedigree. Nail those five checkpoints and you can trust the carbon maths. Skip them and you might as well compare karaoke scores.
Frequently Asked Questions
Can I compare an EN 15804 A2 EPD with an older A1 EPD?
Not directly. The underlying characterisation factors and even some impact categories changed. You would need to recalculate or request an A2-formatted addendum from the manufacturer.
Is cradle-to-gate data ever comparable across products?
Only if both EPDs use identical system boundaries (A1–A3) and the same PCR rules. Even then, downstream impacts like use-phase energy can dominate total carbon, so cradle-to-gate tells just part of the story.
What if two EPDs come from different program operators?
Check for a Mutual Recognition Agreement between the operators. If none exists, assume that additional review of PCR alignment, verifier accreditation, and data quality is necessary before drawing conclusions.
How often do comparability rules change?
Updates hinge on revisions to ISO 14025, EN 15804, or local databases like NMD. Expect meaningful changes every three to five years, with transitional overlap periods.
About the Author
With a passion for sustainability and a love for complex data, Toby helps manufacturers efficiently collect data from across their organization and get their EPDs done right. He’s especially interested in how AI can support human expertise, helping R&D and factory teams work faster, smarter, and with less friction.
